|Publication number||US5515319 A|
|Application number||US 08/135,935|
|Publication date||May 7, 1996|
|Filing date||Oct 12, 1993|
|Priority date||Oct 12, 1993|
|Publication number||08135935, 135935, US 5515319 A, US 5515319A, US-A-5515319, US5515319 A, US5515319A|
|Inventors||Michael C. Smayling, Iano D'Arrigo, Giovanni Santin, Georges Falessi, Mousumi Bhat|
|Original Assignee||Texas Instruments Incorporated|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (11), Referenced by (66), Classifications (27), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This invention generally relates to semiconductor devices and specifically to a non-volatile memory cell as well as a level shifter which may be used together.
A read-only memory (ROM) is a type of memory in which data can be permanently stored, for example, by blowing metallic links during programming thereof. Such type of memory can be conventionally accessed at specific locations to read the programmed contents thereof. A much more versatile type of ROM is the electrically programmable read-only memory (EPROM) which can be electrically programmed to store data. Some versions of these devices store data permanently; others, however, can be erased with ultraviolet light or an electrical current.
The construction of an EPROM cell is similar to that of a field effect transistor, but additionally includes a floating gate between the gate conductor and the conduction channel of the transistor. During programming to store a desired logic state, electrons flowing in the conduction channel are attracted by and trapped in the floating gate. This condition increases the threshold voltage of the transistor, thereby rendering it cutoff in response to normal read operation voltages. Hence, during reading of a transistor cell so programmed, the transistor will remain nonconductive and thus represent a high impedance between the source and drain. On the other hand, cells which have not been programmed will remain conductive in response to normal read operation voltages. In this manner, the memory can be programmed to binary logic states. The memory can also be utilized for analog application where the amount of charge stored on the floating gate is varied with more than just two discrete levels.
Other objects and advantages will be obvious, and will in part appear hereinafter and will be accomplished by the present invention which provides a method and apparatus for a non-volatile memory cell.
A non-volatile memory cell is disclosed herein. The cell is formed in a first semiconductor region of a first conductivity type. A second semiconductor region of a second conductivity type formed over the first semiconductor region. A third semiconductor region of the first conductivity type formed over the second semiconductor region. In the preferred embodiment, the second and third regions are well regions formed within the first region. Other regions such as epitaxially grown layers can also be used. First and second source/drain regions are formed within the third semiconductor region. These second source/drain regions are separated by a channel region. A floating gate overlies at least a portion of the channel region while a control gate overlies the floating gate.
One advantage of the invention is that the source/drain diffusions can be driven with both positive and negative voltages. These lower voltages are both easier to generate and control and less likely to disturb adjacent cells during write and erase operations.
The above features of the present invention will be more clearly understood from consideration of the following descriptions in connection with accompanying drawings in which:
FIG. 1 is a cross-section of a floating gate transistor;
FIG. 2a is a cross-section of a floating gate transistor which is formed within well regions;
FIG. 2b is a cross-section of a floating gate transistor which is formed within semiconductor layers;
FIGS. 3a and 3b are schematic diagrams of arrays of non-volatile memory cells;
FIG. 4 is a cross-section of a prior art cell to illustrate how the writing method of the present invention would be inoperative;
FIGS. 5a and 5b are cross-sections of alternate embodiment floating gate transistors;
FIG. 6 is a cross-section of a level shifter circuit;
FIG. 7 is a schematic diagram of a level shifter circuit;
FIG. 8 is a block diagram using the circuit of FIG. 7 for selecting and deselecting voltages higher than the supply voltage or lower than the reference voltage;
FIG. 9 is a schematic of an array which utilizes a level shifter circuit;
FIG. 10 is a schematic of an alternate embodiment array which utilizes a level shifter circuit;
FIG. 11 is a cross-section of an alternate embodiment floating gate transistor;
FIG. 12 is a schematic of an array of cells such as the one illustrated in FIG. 11;
FIGS. 13a-13f are cross-sections of a floating gate transistor at various stages in the fabrication process;
FIG. 14 is a block diagram of single chip circuit which utilizes a number of simultaneously formed devices;
FIGS. 15a-15d are cross-sections of a simultaneously formed floating gate transistor and vertical bipolar transistor at various stages in the fabrication process;
FIGS. 16a-16d are cross-sections of a simultaneously formed floating gate transistor and extended drain transistor at various stages in the fabrication process; and
FIG. 17 is a cross-section of a floating gate transistor and a CMOS transistor pair which were formed simultaneously;
FIG. 18 is a cross-section of a floating gate transistor and a NMOS transistor pair which were formed simultaneously;
FIG. 19 is a cross section of a floating gate transistor and a DMOS transistor pair which were formed simultaneously.
The making and use of the presently preferred embodiments are discussed below in detail. However, it should be appreciated that the present invention provides many applicable inventive concepts which can be embodied in a wide variety of specific contexts. The specific embodiments discussed are merely illustrative of specific ways to make and use the invention, and do not limit the scope of the invention.
The following is a description of the structure and method of the present invention. The preferred embodiment will be described first followed by a description of modifications. The method of using the invention will then be described.
Referring first to FIG. 1, a preferred embodiment non-volatile memory cell 10 is illustrated. A first semiconductor region 12 is provided. In the preferred embodiment, semiconductor region 12 comprises p-doped silicon for an n-channel memory device. However, if p-channel memory devices are desired, each of the regions could be formed of the opposite conductivity type than illustrated in FIG. 1.
The semiconductor region 12 may comprise a semiconductor substrate. Alternatively, semiconductor region 12 may comprise a semiconductor region 12 formed on another region (not shown). The other region (not shown) may be a semiconductor layer or alternatively may be an insulating layer, for example, if silicon-on-insulator (SOI) technologies are used. In another embodiment, semiconductor region 12 may comprise a well region formed within a semiconductor region. The preferred semiconductor material is silicon but other materials such as germanium or gallium arsenide may also be used.
A second semiconductor region 14 is formed above semiconductor region 12. Semiconductor region 14 is doped to the opposite conductivity type of semiconductor region 12. For example, if semiconductor region 12 comprises p-type semiconductor, then semiconductor region 14 comprises an n-type semiconductor.
A third semiconductor region 16 is formed above semiconductor region 14. As before, semiconductor region 16 is doped to the opposite conductivity type as semiconductor region 14. In a preferred embodiment, the semiconductor region 14 comprises a well region formed within semiconductor region 12, and the semiconductor region 16 comprises a well region formed with semiconductor region 14.
First and second source/drain regions 18 and 20 are formed within semiconductor region 16. In a preferred embodiment, source/drain regions 18 and 20 comprise relatively highly doped n-type regions. However, if a p-channel device is desired, the source/drain regions 18 and 20 would comprise p-type regions. The source/drain regions 18 and 20 are separated by a channel region 22. Typically, the channel region 22 is doped to a conductivity type opposite the source/drain regions 18 and 20.
A gate dielectric layer 24 is formed over the channel region 22. The gate dielectric layer 24 may comprise an oxide layer such as silicon dioxide, a nitride layer such as silicon nitride or a combination of both such as an oxide-nitride (ON) or nitride-oxide (NO) layer or an oxide-nitride-oxide (ONO) layer.
A floating gate 26 is formed over at least a portion of channel region 22 and is insulated therefrom by the gate dielectric 24. As illustrated in FIG. 1, the floating gate 26 may extend the length of the entire channel region 22. In other words, the floating gate 26 may extend from the source 18 to the drain 20. In other embodiments, the floating gate 26 may extend over only a portion of the channel region 22. One known structure where the floating gate 26 extends over only a portion of the channel is a split gate structure. The floating gate 26 comprises a conductive material such as doped polysilicon or a metal such as aluminum.
A control gate 30 is formed over the floating gate 26 and insulated therefrom. The control gate 30 comprises a conductive material such as doped polysilicon or a metal such as aluminum. The control gate 30 may also comprise a silicide such as titanium silicide or tungsten silicide. In typical application, the control gate 30 is coupled to a word line in an array of cells.
Also illustrated in FIG. 1 are sidewall insulating regions 32. The sidewall insulating regions 32 may comprise an oxide such as silicon dioxide or a nitride such as silicon nitride.
In a preferred embodiment, the semiconductor region 14 comprises a well (or tub or tank) region formed within semiconductor region 12, and the semiconductor region 16 comprises a well region formed within semiconductor region 14. An example of this type of structure is illustrated in FIG. 2a.
Two non-volatile memory cells 10a and 10b are illustrated in the structure of FIG. 2a. In this example, cell 10a and cell 10b share a common source/drain region 19. An advantage of forming more than one memory cell in each well 16 is a reduction of surface area. As will be discussed in more detail below, the cells 10a and 10b can be erased together.
An advantage of using well structures as illustrated in FIG. 2a is that electrical contacts can be made to well regions from the top surface of the device. For example, contact regions 15 and 17 are illustrated in FIG. 11. In addition, contact can be made to the first region 12 (as illustrated with contact region 13 in FIG. 11).
An alternative embodiment cell is illustrated in FIG. 2b. In this embodiment, the semiconductor region 14 comprises a semiconductor layer formed over semiconductor region 12 and the semiconductor region 16 comprises a semiconductor layer formed over semiconductor layer 14. These layers 14 and 16 (as well as layer 12) can comprise epitaxially grown semiconductor layers. Alternatively, the layers 14 and 16 may comprise polycrystalline semiconductor layers if the performance requirements allow this material.
In flash EEPROM applications, each of the cells formed in a common well 16 can be erased at the same time. Therefore, if a given application requires the ability to individually erase cells (i.e., erase each bit individually), only a single cell would be formed in each well 16. On the other end of the spectrum, the entire array can be put in a single well. In this embodiment, the entire array would be erased together. In a preferred embodiment, the array is segmented into a series of blocks of cells and each block is formed in a well. One example of an embodiment where this method would be useful is in a disk replacement non-volatile memory device. In this example, each sector (e.g., 512 bytes) could comprise a block and accordingly be formed in a single well.
As previously discussed, the memory cells of the present invention are typically formed in a memory array. A schematic diagram of an array 1 of only twelve cells is illustrated in FIG. 3a. Typical arrays include many more memory cells. For example, it is not uncommon to have arrays of 16K cells or more.
Each of the cells C00 -C23 in the array i may be formed in a separate well or all the cells may be formed in a single well. Typically, the number of cells formed in each well depends on the desired performance (e.g., write or erase time) and is not a functional issue.
In a typical embodiment, the array 1 of non-volatile memory cells is formed in rows and columns. Each of the cells in a column (e.g., C01 -C21) has a first source/drain coupled to a first bitline (e.g., BL1) and a second source/drain coupled to a second bitline (e.g., BL2). Also, each of the cells in a row (e.g., C10 -C13) has a gate coupled to a wordline (e.g., WL1). Each cell may be comprise any of the embodiment cells described herein or variations thereof.
There are several ways to configure the cell within array 1. Each non-volatile memory cell may have the source/drain regions formed in a separate first well region. In this embodiment, there would be the same number of well regions as there are cells. In another embodiment, the source/drain regions of each cell in a column of cells are formed in a single well region. In this embodiment, the number of well regions would be the same as the number of columns. In yet another embodiment, the source/drain regions of each cell in at least two adjacent columns of cells are formed in a single well region. As an example, the cells C00, C01, C10, C11, C20, and C21 may be formed in one well and the cells C02, C03, C12, C13, C22, and C23 may be formed in a second well. (It should be noted, in this example, the cells in the column of C01 would not share bitline BL2 with the cells in the column C02 as illustrated in FIG. 3a.) Any number of cells can be included in each well. In the extreme case, the source/drain regions of each cell in the entire array are formed in a single well region.
Cells in adjacent columns may or may not share source/drain regions. If each cell has its own source/drain regions, the schematic diagram of the array would be as illustrated in FIG. 3b. An example of a pair of cells built with individual source/drain regions is illustrated in FIG. 11. One of the advantages of the memory cells of the present invention is that they may be programmed with both positive and negative voltages. Table 1 illustrates the voltage levels applied to the bitlines (BL0 -BL4) and word lines (WL0 -WL2) of the array shown in FIG. 3a.
TABLE 1__________________________________________________________________________MODE WL0 WL1 WL2 BL0 BL1 BL2 BL3 BL4__________________________________________________________________________Write 1 0V +HV 0V 0V 0V 0V 0V 0VC11 2 0V 0V 0V 0V 0V -HV 0V 0V 3 0V +MV 0V 0V 0V -MV 0V 0VFlash 1 +HV +HV +HV 0V 0V 0V 0V 0VWrite 2 0V 0V 0V -HV -HV -HV -HV -HV 3 +MV +MV +MV -MV -MV -MV -MV -MVRead C11 -Vr +Vr -Vr Float 0V Vs Float FloatErase 1 0V 0V 0V 0V 0V +HV 0V 0VC11 2 0V -HV 0V 0V 0V 0V 0V 0V 3 0V -MV 0V 0V 0V +MV 0V 0VFlash 1 0V 0V 0V +HV +HV +HV +HV +HVErase 2 -HV -HV -HV 0V 0V 0V 0V 0V 3 -MV -MV -MV +MV +MV +MV +MV +MV__________________________________________________________________________ Note: HV = High Voltage (For example, 18 V) MV = Medium Voltage (For example, 9 V) Vr = Word line read voltage (For example, 3 V) Vs = Bit line sense voltage (For example, 1.2 V)
The capability of programming (i.e., writing and erasing) the array with both positive and negative voltage comes from the fact that each well 16 is junction (or oxide) isolated from the region 12 and therefore diffusions (i.e., source/drain regions) can be biased negatively without forward biasing the diffusion/substrate junction.
There are several advantages of the programming method of the present invention. In a standard flash EEPROM cell, the cell is erased by applying zero volts to the control gate 30 (e.g., the wordline) and a positive voltage HV to the source (e.g., the bitline). In other words, the voltage difference between the control gate and source is HV volts (i.e., HV-0=HV). An example of this technique is shown in the first erase example (either flash or cell C11) of Table 1. However, since more than one cell in each array share source/drain regions, the application of the positive voltage to a source region of one cell can adversely effect other cells in the array.
A similar technique is used to write to the cell as illustrated in the first write example of Table 1. In this case, a high voltage +HV is applied to the gate 30 of the cell(s) to be written to and a zero voltage applied to the source (bitline). Although they exhibit some problems, each of these first erasing or writing techniques can be utilized with the cell of the present invention.
An improved erase method is described in U.S. Pat. No. 4,742,492 ('492) issued May 3, 1988, incorporated herein by reference. In the '492 patent, a negative voltage -HV is applied to the control gate 30 and a zero voltage is applied to the source. Here, the voltage difference between control gate and source is still HV volts (i.e., 0-(-HV)=HV). This method provides an advantage that it allows using low voltage isolation within a memory array. This erase technique is shown on the second row of the erase examples in Table 1.
Several disadvantages, however, exist with the previous methods. In typical applications, the voltage V (or -V) is provided by a charge pump. In a typical cell, the voltage V will be about 17 to 18 volts. However, as known in the art, charge pumps tend to become less efficient when higher voltages are needed. A typical charge pump is built with stages. The higher the voltage required, the more stages will be needed. The later stages will be less efficient. Therefore, it is desirable to have an array which requires a lower voltage to be supplied by the charge pump.
The '492 patent also teaches an erase technique where a negative voltage of -MV is applied to the control gate 30 (e.g., wordline) and a positive voltage +MV is applied to the source (e.g., bitline). In this case, the voltage MV is half the voltage HV. Therefore, once again, the voltage difference between control gate and source is still HV volts (i.e., +MV-(-MV)=HV). This method provides an advantage that the circuitry only needs to generate voltage levels which are half as high as previously required. This can be a great advantage, for example, if charge pumps are being used to generate the programming voltages.
The techniques taught in the '492 patent, however, cannot be extended to writing to a cell. FIG. 4 illustrates a prior art cell. In a typical application, the substrate 12 is grounded. If a voltage which is less than the substrate voltage is applied to the source/drain 20, the source/drain-substrate junction will be forward biased and current would flow into the substrate 12 rather than towards the channel 22 as desired.
In the present invention, on the other hand, the source/drain is isolated from the substrate by the intervening layers 14 and 16 of semiconductor. Since the potential of each of these regions can be controlled (via contact) the undesired diode effect can be eliminated.
The second and third lines within the write examples of Table 1 illustrate this writing technique. In the second example zero volts are applied to the gate 30 while a negative voltage -HV is applied to the source/drain 20. In the third example, the positive voltage +MV is applied to the gate 30 while the negative voltage -MV is applied to source/drain 20. The magnitudes of the voltages -MV and +MV may be equal but do not have to be. In other words, the voltages can be any value so long as the difference between the two is HV volts.
FIGS. 5a and 5b illustrate an alternate cell which can be programmed using any of the methods of the present invention. Referring to FIG. 5a, an insulating layer 41 is formed on the surface of semiconductor region 12. The region 12 can be any of the regions previously described. The insulating layer 41 may comprise an oxide, a nitride or a sapphire layer.
In the embodiment of FIG. 5a, the memory cells are formed in semiconductor region 16. In an SOI (silicon on insulator) embodiment, the semiconductor region 16 may comprise a mesa. As described before, one, some or all the cells may be formed in a single mesa. The mesas are isolated from one another by trench 21 which may be filled with an insulating material.
An alternate embodiment is illustrated in FIG. 5b. In this embodiment, adjacent cells are separated by field insulating regions 21 rather than trenches. Other isolation methods can also be used.
The insulating layer 41 may be formed over the entire chip or alternatively may be formed only where needed for the non-volatile memory array. The semiconductor region 16 may be formed by epitaxially depositing a semiconductor layer such as single crystal silicon. Alternatively, polycrystalline silicon can be used as well as other semiconductors such as germanium or gallium arsenide.
It should also be noted that the semiconductor region 14 may comprise a semiconductor layer (e.g., an epitaxially grown layer) while the semiconductor region 16 comprises a well region formed within the layer 14.
In one embodiment of the present invention, the array is coupled to a level shifter with the capability of selecting both positive and negative voltages. One such level shifter will be described next. A circuit which utilizes this level shifter in an EEPROM array will then be described. A level shifter circuit for integrated circuits is also discussed in U.S. Pat. No. 5,157,281 issued Oct. 21, 1992, incorporated herein by reference.
MOS technology allows the generation of negative voltage with respect to circuit ground using a voltage multiplying technique. In the conventional CMOS process, the charge pump circuit uses only p-channel devices due to the necessity to insulate the generated negative voltage from the substrate, which is tied to the circuit ground. Also, no n-channel transistors can be used in the negative voltage path since the N+/P- junction between transistor source or drain and IC substrate would be biased in direct conduction thus shorting the negative voltage to VSS This fact limits negative voltage charge pump application only where a bulk type of negative biasing is needed.
Some circuits have been developed in order to switch a negative voltage, such as a memory row decoder, for example, but their implementation is cumbersome or very difficult to be realized in practice. Some examples include: p-channel depletion transistors, circuits with one negative charge pump for each cell that needs to be independently selected, silicon on insulator circuits, and p-channel only circuitry which requires a more negative voltage to control the switching.
The present invention details a method of building devices with well (or tank or tub) regions. In one example, the wells are left with n type impurities.
Availability of deep n well process used for implementing high power MOS transistors, makes it possible now to design a new type of CMOS inverter-level shifter, able to work and select voltages both higher than VSS and lower than VSS. A cross sectional diagram of a preferred embodiment CMOS shifter 40 is illustrated in FIG. 6.
Referring now to FIG. 6, a semiconductor region 42 is provided. If the level shifter 40 is used in conjunction with an EEPROM cell 10 (as shown in FIG. 1 or 2), the region 42 will be the same as the region 12. Accordingly, semiconductor region 42 has the same characteristics and alternatives as described previously with respect to semiconductor region 12.
The following discussion assumes that the semiconductor region 42 is doped with p-type impurities. However, this is not a requirement. If an n-type region is available, each of the conductivity types would be reversed.
A first well region 44 is formed within semiconductor region 42. The first well region 44 is doped with n-type impurities. A body contact region 46 of more highly doped semiconductor material may be included in well 44 to facilitate electrical contact to the well region 44.
First and second p-channel source/drain regions 48 and 50 are also formed in well region 44. The source/drain regions 48 and 50 are separated by a channel region 52. Gate 54 overlies channel region 52 and is insulated therefrom. The source/drain regions 48 and 50 along with channel 52 and gate 54 form p-channel transistor P1.
Second n-well region 56 is also formed in semiconductor region 42 and is spaced from the first n-well 44. P-well 58 is formed within the second n-well 56. An n-well contact region 60 may also be formed with n-well 56 and a p-well contact region 62 may be formed in p-well 58.
First and second n-channel source/drain regions 64 and 66 are formed in p-well 58. The source/drain regions 64 and 66 are separated by a channel region 68. A gate 70 overlies the channel region 68 and is insulated therefrom. The source/drain regions 64 and 66 along with channel 68 and gate 70 from n-channel transistor N1.
With this configuration, the p-well 58 in which the n-channel device N1 is built, is insulated from the substrate 42 by the deep n-well 56. The n-well 56 may be coupled to voltage node VDD. The voltage Vn can be coupled to source/drain region 66 and also contact region 62. The voltage Vn can be lower than VSS, thus making possible the selection of a negative voltage. The voltage Vn may be provided by a charge pump (not shown). From the p-channel side, the n-well region 44 can be raised higher than VDD, thus making possible to select positive voltages higher than VDD.
If the signals driving gate 54 and gate 70 are at CMOS levels, typically between VSS and VDD, some form of feedback is needed to be able to turn off the p-channel or the n-channel devices. The final form of the circuit may be as shown in FIG. 7.
Referring now to FIG. 7, p-channel transistor P1 and n-channel transistor N1 are provided as described with respect to FIG. 6. Also provided are p-channel feedback transistor P2 and n-channel feedback transistor N2.
The gate 72 of the p-channel feedback transistor P2 is coupled to the output node OUT. The source 74 of the transistor P2 is coupled to the gate 54 of transistor P1 at node PGATE, and the drain 76 is coupled to the voltage node VP. The transistor substrate (which preferably comprises an n-doped well, possibly n-well 44) is also coupled to the voltage node VP.
Similarly, the gate 78 of the n-channel feedback transistor N2 is coupled to the output node OUT. The source 80 is coupled to the gate 70 of the transistor N1 at node NGATE and the drain 82 is coupled to the voltage node VN. The transistor substrate (which may comprise a p-doped well formed in an n-doped well, possibly p-well 58) is also coupled to the voltage node VN.
Access transistors N3 and P3 are also illustrated in FIG. 7. The drain 84 of access transistor N3 is coupled to the node PGATE while the source 86 is coupled to p-side input INP. The gate 88 is coupled to a voltage node such as voltage node VDD. Likewise, the drain 90 of access transistor P3 is coupled to the node NGATE while the source 92 is coupled to n-side input INN. The gate 94 is coupled to a voltage node such as voltage node VSS.
The operation of the circuit will now be explained.
When voltage VP is equal to voltage VDD and voltage VN is equal to VSS, this circuit acts like a CMOS inverter. When INP and INN are at VDD, P1 is off, N1 is on and the output is VSS. The feedback transistor P2 will restore the full VDD voltage level on node PGATE.
Likewise, when INP and INN are at voltage level VSS, P1 is on and N1 is off so the output node OUT is at voltage level VDD. The feedback transistor N2 will restore full VSS voltage level on node NGATE to ensure that transistor N1 is off.
If a voltage higher than VDD needs to be selected on node OUT, VSS level is applied to nodes INP and INN so that OUT will go to the VDD voltage level as explained above. When the switching is complete, the voltage VP is raised over VDD and as a result, the node OUT will follow VP. In this manner, the node OUT will be greater than the level VDD. If a voltage lower than VSS needs to be selected to OUT, VDD will be applied to INP and INN. When the switching is completed and OUT is at VSS, VN can be lowered from VSS and OUT will follow.
When a voltage higher than VDD or lower than VSS needs to be deselected, operations are explained with reference to FIG. 8. Circuits A and B are each circuits as described in FIG. 7.
The operation of the circuit if FIG. 8 will now be described.
If a voltage higher than VDD needs to be selected on OUTA and deselected from OUTB, INPA and INNA are tied to VSS and the circuit A behaves as explained in the previous section. INPB and INNB are tied to VDD and circuit B will behave as follows: Referring to FIG. 7, if VN is equal to VSS, then N1 will turn on so that OUTB will go to VSS. Feedback P2 turns on and raises the node PGATE to voltage level VP thus turning P1 off so that the charge pump driving VP will not be loaded by AC current. Transistor N3 will turn off so VP will be insulated from the circuit driving INPB and INNB.
On the other hand, if a voltage lower than VSS needs to be selected on OUTA and deselected on OUTB, INPA and INNA are tied to VDD and the circuit A behaves as explained on the previous section. INPB and INNB are tied to VSS and circuit B will behave as follows: Referring to FIG. 7, if VP is equal to VDD, then P1 turns on so that OUTB will go to VDD. Feedback transistor N2 turns on and NGATE is lowered to VN that N1 is turned off and the charge pump driving VN is not loaded by AC current. Access transistor P3 will turn off so VN will be insulated from the circuit driving INPB or INNB.
The case in which VP is greater than VDD and VN is less than VSS at the same time is an extension of the two previous cases.
In conclusion, a circuit able to select and deselect voltages higher than VDD and lower than VSS in a CMOS integrated circuit which substrate is biased to VSS and power supply is VDD, has been invented.
The novel circuit can be applied to EEPROM memories. A desirable feature of an EEPROM memory is the ability to alter the status of any single bit from erased to programmed without disturbing the status of the remaining bits in the memory array. But this ability is paid for by a cell size about five times larger than the size of an EPROM cell.
A flash EEPROM is a memory which has a cell size only slightly larger than EPROM and uses the same electrical approach to change status of a single bit. That is, electrons tunnel through a very thin layer of insulating material. For example, this so-called gate oxide may comprise 100 Å of silicon dioxide. But the problem of Flash EEPROM approach is that the memory bits can only be singularly programmed and the erase is a bulk operation. That is, all the bits of the memory are erased at the same time.
This problem is tied to the fact that to erase a bit, a negative voltage is used on the gate side while a positive voltage is applied to the source. This configuration forces the electrons stored on the floating gate to tunnel back and leave a positive charge that is the erase status of the cell. It is the inability to select a negative voltage on a conventional CMOS integrated circuit that creates a root problem.
A possible partial solution of the problem has been to divide the memory array in sections singularly erasable using one a negative charge pump for each sector. However, this architecture becomes impractical if a high grade of operability is desired.
Using the CMOS level shifter with the capability of selecting both positive and negative voltage, the invention of which is described with respect to FIGS. 7 and 8, a novel approach to realize an EEPROM memory is possible.
A first method will be described with reference to FIG. 9 along with Table 2. FIG. 9 illustrates an exemplary four cell array of flash EEPROM cells arranged in two rows and two columns. A first level shifter 100 has an output WLA coupled to the wordline of the first row. In other words, the output WLA is coupled to the gates of the EEPROM cells A and B. Similarly, second level shifter 102 has an output WLB coupled to the wordline of the second row which includes cells B and C. The level shifters 100 and 102 are preferably formed as described above with reference to FIGS. 6 and 7.
Table 2 provides an example of voltages used to program and erase the cell A in the 4-cell array shown in FIG. 9.
TABLE 2__________________________________________________________________________ Deep N-WLA WLB DA DB SA SB VP VN INA INB well__________________________________________________________________________Program A -18V +7V Flt Flt 0V +7V +18V +7V 0V 5V +7VErase A -10 V +5V Flt Flt 5V +0V +5V -10V 5V 0V__________________________________________________________________________
If cell A has to be erased, a voltage of -10V on the gate and +5V on the source will create an electric field over the tunnel region. The intensity of this electric field is enough to tunnel back to the source all the electrons present on the floating gate.
To appreciate the value of this aspect of the present invention, the state of the other three cells (B, C and D) comprising the array must be examined. Regarding cell B, if cell A is erased, the status of cell B will not change with -10V on the gate and 0V on the source. If cell A is programmed, the negative voltage on the floating gate will follow to the voltage on the gate so some extent of erase disturb can be expected. That is, its threshold voltage can be reduced. Regarding cell C, the gate voltage will be 5V and source voltage will be 5V so no electric field can exist over the tunnel region and so its status will not be changed. Regarding cell D, the gate voltage will be 5V and source will be 0V. In this case, the electric field over the tunnel region is well below the threshold of 7 Meg V/cm needed to start any electrons tunneling, so no change of this cell status is expected.
In conclusion, this demonstrates that, a novel approach to EEPROM using Flash EEPROM cell structure is possible.
A second and more conservative approach is also possible in the case that the erase disturb of the first method is an undesirable feature. This second approach still makes use of the novel capability to select a negative voltage.
FIG. 10 is an example of a memory array using the second method. Buffer RAM 104 is typically a static RAM cell interfacing with the columns of the array. The function of buffer RAM 104 is to store the data to be written in a row when using the page mode programming. This scheme provides a standard method to write in EEPROM and Flash EEPROM memories.
If cell A only needs to be reprogrammed, the first step will be to read row A and store its content in to the buffer RAM 104.
The second step will be to erase the whole row A by selecting a negative voltage on wordline WLA, while the deselected wordline WLB goes to 5V. All sources lines are tied to 5V as described with the first method, cell A and B are erased while cells C and D, having both gate and source at the same potential of 5V, are not disturbed.
On the third step, the data to be written in cell A is stored in the buffer RAM 104 connected to the column A, while the remaining cells content is unchanged. The fourth step is to write into row A the contents of the buffer RAM 104. At the end, the result is that only the content of cell A has been modified while all the remaining cells of row A and B have their contents unchanged.
This method demonstrates that using the technique of selecting and deselecting negative voltage, a novel architecture to fabricate EEPROM memories with the same small cell size of a Flash EEPROM memory, has been invented.
Referring now to FIGS. 11 and 12, flash EEPROM array with a separate p-well, insulated from the substrate by a deep n-well, is illustrated.
Referring now to FIG. 11, a structure which includes two flash EEPROM cells 10a and 10b formed in a p-well 16 is illustrated. P-well 16 is formed in n-well 14 which in turn is formed in semiconductor region 12. As discussed regarding semiconductor region 12 shown in FIG. 1, the region 12 may be one of many types of semiconductor regions.
Each of the memory devices 10 comprise a source region 18 and a drain region 20. The source/drain regions are formed beneath an insulating region and therefore are sometimes referred to as buried regions. The source/drain regions 18 and 20 are separated by a channel region 22. Formed above a portion of channel region 22 is floating gate 26. Unlike the floating gate structure illustrated in FIGS. 1 and 2, the floating gate 26 of this embodiment extends over only a portion of the channel 22. A control gate (or wordline) 30 is formed over both transistor floating gates 26 and channel regions 22.
Also illustrated in FIG. 11 are field insulating regions 21. The regions 21 are provided to electrically isolate portions of the device. Typically, the field insulating regions 21 comprise an oxide such as thermally grown silicon dioxide.
In addition, each well region may include a more heavily doped contact region. Accordingly, contact region 13 comprises more heavily doped p-type material and is formed in semiconductor region 12. Also, more heavily doped n-type contact region 15 is formed in n-well 14 and more heavily doped p-type contact region 17 is formed in p-well 16.
FIG. 12 illustrates a four cell array formed with two rows and two columns. The exemplary cell A has been labeled to correspond with the structure of FIG. 11.
Table 3 provides an example of specific voltage levels which may be applied to the array of FIG. 12 in order to operate the array. Of course, other values may be used for other specific embodiments.
TABLE 3__________________________________________________________________________WLA WLB AA SA DB SB ARSUB DEPN+__________________________________________________________________________Read A3.5V 0V 0V Fl 0V 1.5V 0 5VWrite A10V 0V Fl 0V Fl -8V -8V 5VErase A-10V 5V Fl 0V Fl 5V 0V 5V__________________________________________________________________________
This embodiment provides a number of advantages. First, the p-well 16 and also the source SB (20) can be charge pumped to -8V. In this way, in order to have 18V of voltage drop over the cell (a typical programming voltage for a flash EEPROM cell), a maximum voltage of only 10V must be generated for the word line. This will greatly alleviate the demand for a high voltage process.
A second advantage of having an insulated p-well 16 for the array is to permit the deselection of an erased one transistor cell. In fact, if the pass gate is omitted from the cell, an eventual over erase results in a depletion status. This phenomena causes the impossibility to deselect the cell using 0V on the deselected word line. But, if the substrate is pumped slightly negative, this will add to the cell threshold voltage, and so the positive dosage present on the floating gate will be no longer able to keep this cell conducting. The conditions for reading the cell are shown in Table 4.
TABLE 4__________________________________________________________________________WLA WLB AA SA DB SB ARSUB DEPN+__________________________________________________________________________Read A3V 0V 0V Fl 0V 1.5V -3V V.sub.DD__________________________________________________________________________
An exemplary fabrication process flow will now be described with respect to FIGS. 13a-13f. It should be noted that variations to the process flow of FIGS. 13a-13f can be included.
Referring now to FIG. 13a, semiconductor region 12 is provided as previously described. A masking layer 150 is formed on the surface of semiconductor region 12 and is patterned using an of a large number of lithography techniques. A first well region 14 is then formed within region 12. Preferably the well 14 is implanted but other doping techniques such as diffusion can also be used. Arsenic and/or phosphorus are the typical n-type dopants but other dopants may be used as well.
Referring next to FIG. 13b, the mask layer 150 is removed and mask 152 is formed. The second well 16 is formed within the region exposed by mask 152. As illustrated in FIG. 13b, the well 16 is formed within well 14. Boron is the typical p-type dopant but other dopants may be used as well.
In an alternate embodiment, both well regions 14 and 16 can be formed with a single mask as disclosed in U.S. Pat. No. 5,242,811 issued Sep. 7, 1993, incorporated herein by reference. In this embodiment, the dopants of well region 14 are chosen so that they have a greater diffusity than the dopants of well region 16. Because of this characteristic, when the device is heated, the region 14 dopants will diffuse farther and region 16 will be formed within region 14.
Referring now to FIG. 13c, field isolating regions 21 are formed. The field isolating regions 21 may be formed using a LOCOS (local oxidation of silicon) and therefore comprise a thermal oxide such as silicon dioxide. Other methods may also be used.
The next step is to form the gate structures of the EEPROM device. In a typical embodiment, a blanket layer of polysilicon is formed over the device. Next, an interlevel dielectric (e.g., oxide, nitride or combination thereof) is formed followed by a second layer of doped polysilicon. The polysilicon may be doped in situ or doped subsequent to deposition. The conductive and dielectric layers are then etched to form the gate structure which includes floating gate 26 and control gate 30. These gate structures are illustrated in FIG. 13d. Other materials could be used for the gate (e.g., aluminum or another metal).
Referring now to FIG. 13e, the more highly doped n-regions are formed. In the preferred embodiment, a mask layer (not shown) is formed over the regions which will not be doped. (Noting it is not typically necessary to form a mask over the field insulating regions 21 although it will not hurt). Source/drain regions 18, 19 and 20 and well contact region 15 are formed in this doping step. As will be discussed in more detail below, other devices may be simultaneously formed. Once again, the doping step preferably comprises an implantation step.
Referring now to FIG. 13f, the more highly doped p-regions are formed. In this case, the p-well contact region 17 and substrate contact region 13 are formed. It should be noted that the fact that the n regions were formed before the p regions is a matter of totally arbitrary choice. Accordingly, the p regions 13 and 17 may have been performed before n-regions 15, 18, 19 and 20.
It should also be noted that a device which includes separate source/drain regions (that is, rather than a shared source/drain region 20) may be formed with minor variation. Also, buried source/drain regions as illustrated in FIG. 9 could also be formed by including buried isolation regions which could be formed before the highly doped regions are formed.
The present invention may be used in a variety of applications. Some specific examples include disk replacement applications, a boot up memory for microprocessors or microcontrollers, and data storage units.
In particular, the present invention is suited for a modular process flow such as is described in U.S. Pat. No. 5,242,811, incorporated herein by reference. This process strategy is sometimes referred to as PRISM. A block diagram of one exemplary PRISM circuit is illustrated in FIG. 14. As shown, the circuit may include some or all of phase lock loops, oscillators, EPROMs, voltage regulators, SCI (serial communication interface), SPI (serial peripheral interface), central processing units, random access memories, real only memories, decoders, direct memory access controllers, and gage drivers as well as EEPROMs as described herein.
Among the advantages provided, with an integrated process flow, that is where particular devices may be omitted if not required, some of the process steps may be omitted. For example, if the particular chip to be fabricated does not require power transistors having buffed drain regions or collectors, then a (n+) buried layer step, an EPI step, and a deep (n+) step may be omitted. Similarly, if the device does not require FAMOS EPROM cells, a FAMOS floating gate step and a EPROM stack step may be omitted. Additions to the process can be made as well for modifications of cells.
In addition to the modular nature of the process, process steps typically at the end of the process flow can be performed at the beginning of the process flow. An advantage of this is that key parameters of the invention, such as channel length, are defined by the diffusion rate of the dopants and not by the placement of the well with respect to a specific structure.
The inventive process provides a unified set of process parameters that can be applied to distinctly different devices. This allows design rules for the different devices to be stored in a library. An integrated circuit designer may select different devices from this library with the assurance that a process is available for building them, and that the devices so selected are compatible with this process. This considerably reduces design time for chips having new designs.
FIGS. 14a-14d and 15a-15d illustrate but two pairs of the number of devices which may be simultaneously formed using the PRISM process. In FIGS. 14a-14d, a vertical bipolar transistor is formed simultaneously with an EEPROM device. In FIGS. 15a-15d, a drain extended transistor is formed simultaneously with an EEPROM device. As will be clear, the vertical bipolar transistor, the drain extended transistor and the EEPROM could all have been formed simultaneously. In addition, other devices such as double diffused MOS (DMOS) transistor and lateral bipolar transistors as well as more standard MOS (n-channel, p+channel and CMOS) can also be simultaneously formed.
Referring first to FIG. 15a, first n-well 14 and second n-well 154 as well as first p-well 16 and second p-well 156 are formed in semiconductor region 12 as previously described. The n-well region 154 will comprise the collector of the transistor while p-well region 156 comprises the base. The field insulating regions 21 and gate structure 26 and 30 are then formed as illustrated in FIG. 15b
Referring now to FIG. 15c, EEPROM source/drain regions 18 and 20 and n-well contact region 15 are formed as previously described. During this step, emitter region 158 and collector contact region 160 are also formed. Likewise, base contact region 162 is formed simultaneously with p-well contact 17. (Possibly as well as substrate contact 13 which is not illustrated). This step is illustrated in FIG. 15d. Once again, the n and p doping steps may be reversed.
As another example, FIGS. 16a-16d illustrate the simultaneous fabrication of an extended drain field effect transistor and an EEPROM cell. First and second n-well regions 14 and 164 and first and second p-well regions 16 and 166 are formed in semiconductor region 12. Field isolating regions 21 are then formed. These steps are illustrated in FIGS. 16a and 16b.
Referring now to FIG. 16c, the gate structure is formed. Transistor gate 158 is formed at the same time as EEPROM floating gate 26. Accordingly, gate 168 may comprise doped polysilicon or a metal such as aluminum.
Referring now to FIG. 16d, the highly doped n and p regions are formed. In this embodiment, the source/drain regions 170 and 172 are formed simultaneously with p-well contact region 17 (and possibly substrate contact 13 which is now shown here).
FIG. 17 illustrates a CMOS device which has been built simultaneously with a EEPROM device. The intervening process steps have not been illustrated since they correspond with the steps already illustrated in FIGS. 13, 15 and 16. The p-channel device of the CMOS transistor pair is formed n-well 174. The p-channel device comprises n-doped source/drain regions 182 and 184 which are separated by a channel region 183. Gate 188 overlies channel 183. The source/drain regions 182 and 184 were formed at the same time as p-contact 17. The gate was formed at the same time as floating gate 26.
The n-channel transistor of the CMOS pair is formed in p-well 176. The n-channel transistor comprises source/drain regions 178 and 180 which are separated by channel 179. Gate 186 overlies the channel region. The source/drain regions 178 and 180 were formed simultaneously with n-regions 15, 18 and 20 while the gate 180 was formed simultaneously with floating gate 26 (as well as p-channel gate 188).
FIG. 18 illustrates a NMOS device which has been built simultaneously with a EEPROM device. Once again, the intervening process steps have not been illustrated since they correspond with the steps already illustrated in FIGS. 13, 15 and 16. The NMOS device of FIG. 18 has been illustrated to specifically include lightly doped drain (LDD) regions 190 and 192. It should be noted that these may also be included in the transistors of FIG. 17 or may be omitted in the NMOS device of FIG. 18.
As a final illustrated example (although other devices could also be incorporated), FIG. 19 illustrates a DMOS device which has been built simultaneously with a EEPROM device. First source/drain region 198 is formed in p-well 196. Second source/drain region 200 is spaced from the first source/drain region 198. The channel of the device is formed within p-well 196. The gate 202 is formed so that it overlies at least a portion of the channel.
It should once again be noted that any combination of the devices described herein can be formed simultaneously. In other words the present invention encompasses forming an EEPROM cell simultaneously with any one, some or all of the devices described herein as well as other devices. The present invention also encompasses combinations of devices which do not include the EEPROM cells.
While this invention has been described with reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. Various modifications and combinations of the illustrative embodiments, as well as other embodiments of the invention, will be apparent to persons skilled in the art upon reference to the description. It is therefore intended that the appended claims encompass any such modifications or embodiments.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4016588 *||Dec 24, 1975||Apr 5, 1977||Nippon Electric Company, Ltd.||Non-volatile semiconductor memory device|
|US4742492 *||Sep 27, 1985||May 3, 1988||Texas Instruments Incorporated||EEPROM memory cell having improved breakdown characteristics and driving circuitry therefor|
|US4888735 *||Apr 28, 1988||Dec 19, 1989||Elite Semiconductor & Systems Int'l., Inc.||ROM cell and array configuration|
|US5016215 *||Mar 12, 1990||May 14, 1991||Texas Instruments Incorporated||High speed EPROM with reverse polarity voltages applied to source and drain regions during reading and writing|
|US5060195 *||Oct 11, 1990||Oct 22, 1991||Texas Instruments Incorporated||Hot electron programmable, tunnel electron erasable contactless EEPROM|
|US5097303 *||Mar 29, 1991||Mar 17, 1992||Fujitsu Limited||On-chip voltage regulator and semiconductor memory device using the same|
|US5157281 *||Jul 12, 1991||Oct 20, 1992||Texas Instruments Incorporated||Level-shifter circuit for integrated circuits|
|US5243559 *||Dec 12, 1991||Sep 7, 1993||Nippon Steel Corporation||Semiconductor memory device|
|US5287318 *||Feb 14, 1992||Feb 15, 1994||Sharp Kabushiki Kaisha||Semiconductor memory|
|US5341342 *||Dec 18, 1992||Aug 23, 1994||National Semiconductor Corporation||Flash memory cell structure|
|US5371706 *||Aug 20, 1992||Dec 6, 1994||Texas Instruments Incorporated||Circuit and method for sensing depletion of memory cells|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US5576995 *||Oct 24, 1995||Nov 19, 1996||Sharp Kabushiki Kaisha||Method for rewriting a flash memory|
|US5732021 *||Jul 19, 1996||Mar 24, 1998||Smayling; Michael C.||Programmable and convertible non-volatile memory array|
|US5761126 *||Feb 7, 1997||Jun 2, 1998||National Semiconductor Corporation||Single-poly EPROM cell that utilizes a reduced programming voltage to program the cell|
|US5793678 *||Apr 5, 1996||Aug 11, 1998||Hitachi, Ltd.||Parellel type nonvolatile semiconductor memory device method of using the same|
|US5920504 *||Sep 12, 1997||Jul 6, 1999||Samsung Electronics Co., Ltd.||Semiconductor memory having embedding lockable cells|
|US5940324 *||Apr 1, 1998||Aug 17, 1999||National Semiconductor Corporation||Single-poly EEPROM cell that is programmable and erasable in a low-voltage environment|
|US5978269 *||Aug 17, 1998||Nov 2, 1999||National Semiconductor Corporation||Apparatus and method for lowering the potential barrier across the source-to-well junction during the programming of non-volatile memory cells|
|US6043123 *||May 27, 1997||Mar 28, 2000||Hyundai Electronics America, Inc.||Triple well flash memory fabrication process|
|US6067251 *||Oct 27, 1998||May 23, 2000||Sharp Kabushiki Kaisha||Non-volatile semiconductor memory device|
|US6330190 *||May 27, 1997||Dec 11, 2001||Hyundai Electronics America||Semiconductor structure for flash memory enabling low operating potentials|
|US6347054||Feb 1, 2000||Feb 12, 2002||Hyundai Electronics America||Method of operating flash memory|
|US6359461 *||Feb 10, 1998||Mar 19, 2002||Advanced Micro Devices, Inc.||Test structure for determining the properties of densely packed transistors|
|US6377490||Oct 3, 2000||Apr 23, 2002||Matsushita Electronics Corporation||Nonvolatile semiconductor memory device and method for driving the same|
|US6406953 *||Apr 1, 1998||Jun 18, 2002||Mosel Vitelic, Inc.||Method for fabricating an integrated circuit with a transistor electrode|
|US6509606||Apr 1, 1998||Jan 21, 2003||National Semiconductor Corporation||Single poly EPROM cell having smaller size and improved data retention compatible with advanced CMOS process|
|US6525970||Oct 12, 2001||Feb 25, 2003||Hyundai Electronics America||Erase method for flash memory|
|US6549458 *||Oct 25, 2001||Apr 15, 2003||Xilinx, Inc.||Non-volatile memory array using gate breakdown structures|
|US6566705 *||Dec 20, 2001||May 20, 2003||Intersil Americas, Inc.||Enhanced EPROM structures with accentuated hot electron generation regions|
|US6621116||Dec 20, 2001||Sep 16, 2003||Michael David Church||Enhanced EPROM structures with accentuated hot electron generation regions|
|US6636441 *||Nov 15, 2001||Oct 21, 2003||Kabushiki Kaisha Toshiba||Semiconductor memory device including non-volatile memory cell array having MOS structure in well region formed on semiconductor substrate|
|US6657893||Jan 22, 2002||Dec 2, 2003||Matsushita Electric Industrial Co., Ltd.||Nonvolatile semiconductor memory device and method for driving the same|
|US6677200 *||Jul 12, 2002||Jan 13, 2004||Samsung Electronics Co., Ltd.||Method of forming non-volatile memory having floating trap type device|
|US6680507 *||Apr 11, 2001||Jan 20, 2004||Advanced Micro Devices||Dual bit isolation scheme for flash memory devices having polysilicon floating gates|
|US6777280||Apr 30, 2002||Aug 17, 2004||Mosel Vitelic, Inc.||Method for fabricating an integrated circuit with a transistor electrode|
|US6888753||Oct 6, 2003||May 3, 2005||Infineon Technologies Ag||Memory cell array comprising individually addressable memory cells and method of making the same|
|US6905929||Oct 28, 2002||Jun 14, 2005||National Semiconductor Corporation||Single poly EPROM cell having smaller size and improved data retention compatible with advanced CMOS process|
|US6921690||Dec 20, 2001||Jul 26, 2005||Intersil Americas Inc.||Method of fabricating enhanced EPROM structures with accentuated hot electron generation regions|
|US6930350 *||Sep 4, 2003||Aug 16, 2005||Micron Technology, Inc.||Contactless uniform-tunneling separate p-well (CUSP) non-volatile memory array architecture, fabrication and operation|
|US7151697||Nov 30, 2004||Dec 19, 2006||Infineon Technologies Ag||Non-volatile semiconductor memory|
|US7176519||Aug 5, 2005||Feb 13, 2007||Infineon Technologies Ag||Memory cell, memory cell arrangement and method for the production of a memory cell|
|US7199422||Dec 6, 2004||Apr 3, 2007||Micron Technology, Inc.||Contactless uniform-tunneling separate P-well (CUSP) non-volatile memory array architecture, fabrication and operation|
|US7504861 *||Jun 21, 2004||Mar 17, 2009||Transpacific Ip, Ltd.||Input stage for mixed-voltage-tolerant buffer with reduced leakage|
|US7544558 *||Mar 13, 2006||Jun 9, 2009||Bcd Semiconductor Manufacturing Limited||Method for integrating DMOS into sub-micron CMOS process|
|US7696557||Feb 15, 2007||Apr 13, 2010||Micron Technology, Inc.||Contactless uniform-tunneling separate p-well (CUSP) non-volatile memory array architecture, fabrication and operation|
|US7948800 *||Jan 13, 2009||May 24, 2011||Panasonic Corporation||Semiconductor memory device and driving method for the same|
|US7969190||Mar 16, 2009||Jun 28, 2011||Che-Hao Chuang||Input stage for mixed-voltage-tolerant buffer with reduced leakage|
|US8320191 *||Mar 14, 2008||Nov 27, 2012||Infineon Technologies Ag||Memory cell arrangement, method for controlling a memory cell, memory array and electronic device|
|US8705278 *||Nov 5, 2010||Apr 22, 2014||Macronix International Co., Ltd.||One-transistor cell semiconductor on insulator random access memory|
|US8736015 *||Sep 27, 2011||May 27, 2014||Taiwan Semiconductor Manufacturing Company, Ltd.||Integrated circuit structure and method of forming the same|
|US9030877||Oct 11, 2012||May 12, 2015||Infineon Technologies Ag||Memory cell arrangement, method for controlling a memory cell, memory array and electronic device|
|US9111985 *||Jan 11, 2007||Aug 18, 2015||Cypress Semiconductor Corporation||Shallow bipolar junction transistor|
|US9136326 *||Jan 16, 2014||Sep 15, 2015||Fuji Electric Co., Ltd.||Semiconductor device with increased ESD resistance and manufacturing method thereof|
|US20030119236 *||Dec 20, 2001||Jun 26, 2003||Church Michael David||Method of fabricating enhanced EPROM structures with accentuated hot electron generation regions|
|US20040071008 *||Sep 4, 2003||Apr 15, 2004||Micron Technology, Inc.||Contactless uniform-tunneling separate p-well (CUSP) non-volatile memory array architecture, fabrication and operation|
|US20040079986 *||Oct 6, 2003||Apr 29, 2004||Roland Kakoschke||Memory cell array comprising individually addressable memory cells and method of making the same|
|US20040224452 *||Jun 3, 2004||Nov 11, 2004||Li-Chun Li||Method for fabricating an integrated circuit with a transistor electrode|
|US20050099846 *||Dec 6, 2004||May 12, 2005||Micron Technology, Inc.||Contactless uniform-tunneling separate P-well (CUSP) non-volatile memory array architecture, fabrication and operation|
|US20050110520 *||Jun 21, 2004||May 26, 2005||Industrial Technology Research Institute.||Input stage for mixed-voltage-tolerant buffer without leakage issue|
|US20050285182 *||Aug 5, 2005||Dec 29, 2005||Franz Schuler||Memory cell, memory cell arrangement and method for the production of a memory cell|
|US20060114724 *||Nov 30, 2004||Jun 1, 2006||Stephan Riedel||Non-volatile semiconductor memory|
|US20070164348 *||Feb 15, 2007||Jul 19, 2007||Micron Technology, Inc.||Contactless uniform-tunneling separate p-well (cusp)non-volatile memory array architecture, fabrication and operation|
|US20070212823 *||Mar 13, 2006||Sep 13, 2007||Bcd Semiconductor Manufacturing Limited||Method for integrating DMOS into sub-micron CMOS process|
|US20090059678 *||Mar 14, 2008||Mar 5, 2009||Robert Strenz||Memory Cell Arrangement, Method for Controlling a Memory Cell, Memory Array and Electronic Device|
|US20090185427 *||Jan 13, 2009||Jul 23, 2009||Keita Takahashi||Semiconductor memory device and driving method for the same|
|US20090195269 *||Mar 16, 2009||Aug 6, 2009||Che-Hao Chuang||Input stage for mixed-voltage-tolerant buffer with reduced leakage|
|US20100117153 *||Nov 7, 2008||May 13, 2010||Honeywell International Inc.||High voltage soi cmos device and method of manufacture|
|US20110095353 *||Nov 5, 2010||Apr 28, 2011||Macronix International Co., Ltd.||One-transistor cell semiconductor on insulator random access memory|
|US20140197491 *||Jan 16, 2014||Jul 17, 2014||Fuji Electric Co., Ltd.||Semiconductor device and manufacturing method thereof|
|USRE37311 *||Dec 2, 1999||Aug 7, 2001||Hitachi, Ltd.||Parallel type nonvolatile semiconductor memory device and method of using the same|
|EP0810667A2 *||May 30, 1997||Dec 3, 1997||Hyundai Electronics America, Inc.||Triple well flash memory cell and fabrication process|
|EP0810667A3 *||May 30, 1997||Dec 10, 1997||Hyundai Electronics America, Inc.||Triple well flash memory cell and fabrication process|
|EP0924768A3 *||Dec 4, 1998||Mar 1, 2000||Matsushita Electronics Corporation||Nonvolatile semiconductor memory device and method for driving the same|
|WO2002082550A2 *||Mar 19, 2002||Oct 17, 2002||Infineon Technologies Ag||Memory cell array and method for the production thereof|
|WO2002082550A3 *||Mar 19, 2002||Mar 20, 2003||Infineon Technologies Ag||Memory cell array and method for the production thereof|
|WO2004070841A2 *||Feb 4, 2004||Aug 19, 2004||Infineon Technologies Ag||Storage cell and method for the production thereof|
|WO2004070841A3 *||Feb 4, 2004||Nov 4, 2004||Infineon Technologies Ag||Storage cell and method for the production thereof|
|U.S. Classification||365/185.27, 365/185.26, 365/185.16, 365/185.06, 365/185.33, 257/E29.306, 257/E21.689, 257/E27.103, 365/185.18, 257/E27.081|
|International Classification||H01L21/8247, G11C16/10, H01L29/788, H01L27/105, H01L27/115|
|Cooperative Classification||H01L27/105, H01L27/11526, H01L29/7885, G11C16/10, H01L27/115, H01L27/11546|
|European Classification||H01L27/115F6P2, H01L27/105, G11C16/10, H01L29/788B6B, H01L27/115, H01L27/115F6|
|Dec 16, 1993||AS||Assignment|
Owner name: TEXAS INSTRUMENTS INCORPORATED
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:D ARRIGO, IANO;REEL/FRAME:006864/0912
Effective date: 19931207
Owner name: TEXAS INSTRUMENTS INCORPORATED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:BHAT, MOUSUMI;REEL/FRAME:006864/0921
Effective date: 19931202
Owner name: TEXAS INSTRUMENTS INCORPORATED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:FALESSI, GEORGES;REEL/FRAME:006864/0918
Effective date: 19931202
Owner name: TEXAS INSTRUMENTS INCORPORATED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SMAYLING, MICHAEL C.;REEL/FRAME:006864/0909
Effective date: 19931202
Owner name: TEXAS INSTRUMENTS INCORPORTED, TEXAS
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:SANTIN, GIOVANNI;REEL/FRAME:006864/0915
Effective date: 19931203
|Oct 1, 1999||FPAY||Fee payment|
Year of fee payment: 4
|Sep 26, 2003||FPAY||Fee payment|
Year of fee payment: 8
|Sep 14, 2007||FPAY||Fee payment|
Year of fee payment: 12